Tip: Solving Kubernetes workers moved to unschedulable nodes

Today, I have seen that my kubernetes cluster was switched to cordon mode, where all nodes suddenly become not supporting pod scheduling.

I thought to share the commands and sample of the screen outputs.

https://kubernetes.io/docs/tasks/administer-cluster/safely-drain-node/

First, I have listed the pods and I found many pods are in a pending state, this means that Kubernetes is not yet assigned the node to a node to run any of these pods.

I ran normal get pod commands as seen below!

Containers are in pending state (not scheduled)

Second, to check what’s happening, I run get nodes command to see, and status was clear that the nodes have been configured to Disable Scheduling.

So I stared with enabling scheduling on nodes, by removing them from current cordon they are in!

I have run again the get pod command to see the progress.

Now containers start to popup (Starting)

I have run get events command to get what is happening behind the scene, it seems that everything goes back to normal, excepts a little DiskPressure happened, this is normal (not having the high-speed disk expected) as there are many pods are scheduled at the same time.